High Intensity

As pointed out earlier, the scintillator trigger functionality is reduced at high intensity, as particles arrive with every bunch. The difference between scintillator and random triggering at a beam intensity of $1.4\,\rm MHz$ is illustrated in fig. . It shows the number of measured hits per strip in the Vienna Milestone module. The last two APV6 chips were not fully functional, so their strips are not included in this plot. Regardless of the trigger type, the wide-spread high intensity beam causes a pedestal of approximately 80 hits in every channel. With the beam trigger, the hits in the area covered by the scintillator are preferredly read out, but other particles contained in the same bunches still produce the same background. At low intensity, the beam profile is the same but reduced by the background level.

Figure: Measured hit profiles at low intensity with scintillator trigger (blue) and at high intensity with scintillator (black) and random (red) triggers.

The absolute particle flux is proportional to the measured hit profile with random trigger. However, it is difficult to state the proportionality factor. First of all, our high intensity measurements had an enormous dead time. Further triggers were blocked until the computer finished reading out the previous event data from the ADCs, which took about $5\,\rm ms$. With particle bunches arriving every $20\,\rm ns$, the sensitive time window is only $4\,\rm ppm$. An uncertainty of the actual readout time propagates to this fraction. Moreover, the hit discrimination strongly differs between peak and deconvolution modes resulting in different time over threshold windows, which also depend on the signal amplitude and the timing with respect to the APV clock. Thus, we relied on the scintillator for the beam intensity measurements.

Figure: Measured distribution of the number of hits per random trigger at high intensity in peak mode, fitted by a Poisson distribution.

The statistical character of the high intensity PSI pion beam was also verified by the distribution of the number of hits per time slot. For better statistics, this measurement has been performed in peak mode, where the time window is open longer for particle hits than in deconvolution mode. Fig.  shows the multiplicity distribution obtained from the Vienna Milestone module at high beam intensity. The good agreement with the Poisson distribution demonstrates the highly random nature of the beam.

Apart from the above investigations, normal runs have been made during a period of more than 24 hours at high intensity. The module performance remained unchanged during these runs compared to low beam intensity.

Since the detector currents were continuously monitored, the leakage current increase could be observed as a function of total dose. This is best demonstrated by the Vienna APV25 module, which had an extremely low initial leakage current. Fig.  shows the current development of the V25 module at $-10^{\circ}\,\rm C$ at high beam intensity.

Figure: Detector current of the V25 module during the high intensity period.

Initially, the current is very low, but jumps by approximately $0.5\,\rm\mu A$ as soon as the high intensity beam is turned on. This is caused by the large number of carriers generated within the detector by the crossing particles, therefore called ``beam induced current''. Although the current contribution of a single particle is very small and of short duration, as demonstrated by the detector simulation discussed in section , p. , the huge number of particles over the large area of the two sensors result in a significant DC current which has to be delivered by the power supply.

The slope of the current in fig.  corresponds to the increase of leakage current caused by radiation defects (see section , p. ). A current related damage rate of $\alpha\approx 8\cdot 10^{-17}\,\rm A/cm$ at room temperature has been extracted from this measurement, which agrees with values given by the RD48 collaboration [22].

Occasionally, the beam went off for a few seconds, resulting in a measurement of the pure leakage current without the beam induced component. Joining these points would result in a parallel line of the same slope but at lower level. The current peak in the center was resulting from a power failure of the slow control computer, thus switching off the cooling for half an hour. During this period, the temperature in the cooling box warmed up to $0^{\circ}\,\rm C$, resulting in a dark current increase by a factor of 2.7 according to eq. , p. , which was actually observed. After restoring the operating temperature, the current curve continues unaffected. Later, the beam intensity was slightly reduced for about two hours, resulting in a lower beam induced current.